Disputes over the reasons for Sweden's economic success: Nima Sanandaji and his critics

Sanandaji recalls that long before the 1970s and 1980s, when the growth of the welfare state reached its peak, Sweden had achieved a very high level of Gross Domestic Product (GDP) per capita. He writes that between 1870 and 1936 the economic growth rate in Sweden was the highest among the industrialized countries, while between 1936 and 2008 Sweden was ranked thirteenth in this respect [Sanandaji, 2016a, p. 15].

In his opinion, the driving force behind this early economic expansion was mainly due to the huge amount of social capital accumulated in Sweden (a high work ethic that affirms reliable and hard work and mutual trust that fosters cooperation between people). This economic development was also supported by the Swedes’ conviction of the individual responsibility of people for their actions and their belief in egalitarian ideals.

The reason for Sweden's economic success was also the rapid development of the market, freed up by reforms in the second half of the 19th century. Its unrestricted operation enforced economic efficiency [Sanandaji, 2016a, pp. 32–33; Sanandaji, 2016b, pp. 86–89]. The development of the Swedish economy has also been affected by the freedom of trade and the construction of the railway network. The creation of a modern banking sector and the privatization of forests were of great importance [Sanandaji, 2016b, pp. 87, 90].

According to Sanandaji, the huge welfare state has become the cause not of success, but of failure in Swedish society. Built in the 19th century and the first half of the 20th century, prosperity enabled Swedish social democrats to create this leviathan. The funds for its maintenance have been provided by increasingly higher taxes. In 1965, the share of state revenues from taxes in Swedish GDP, previously similar to that of America and the United Kingdom, was 31% and growing rapidly, reaching 39% in 1975 and 45% in 1985 (compared to 24%, 25%, and 25% in the United States and 29%, 34%, and 36% in the United Kingdom respectively) [Sanandaji, 2016a, p. 37].

The aim of the social-democratic economic policy included, among other things, a reduction in the role of private property in the economy. From 1984, the state began to take over part of the profits of Swedish companies, thus financing the buy-back of their shares, which were then transferred to trade union-controlled “workers’ funds,”; this was done gradually to bring socialism to the economy [Sanandaji, 2016a, p. 43; Sanandaji, 2016b, pp. 91–93].

Thus, Sanandaji claims, such experiments were only possible because they were carried out in a prosperous and rapidly developing country with large social capital resources. It is precisely because of this that the weakening of incentives for economic efficiency, caused by the excessive growth of the welfare state and attempts to collectivize private enterprises, only slowed down the growth of the Swedish economy after 30 years, rather than much earlier. Sanandaji indicates, among other things, that in the 1970s very few new businesses and new jobs were created in Sweden [Sanandaji, 2016a, pp. 15–16]. As a result, in 1975–1995 Sweden dropped from 4th to 13th place in the Organisation for Economic Co-operation and Development (OECD) countries’ wealth ranking (in terms of GDP per capita) [Sanandaji, 2016a, p. 45].

The slowdown in economic growth was accompanied by the disappearance of social capital, which was destroyed by an ideologized and ill-constructed welfare state. Exorbitant taxes not only damaged entrepreneurship but also encouraged tax fraud and corruption. Excessive benefits made recipients dependent, resulting in job avoidance and claimant attitudes [Sanandaji, 2016a, pp. 83–90; 92–98; Sanandaji, 2016b, pp. 131–135].

In support of his words, the author of Debunking Utopia quotes, among others, the results of the World Value Survey, documenting the phenomenon of the destruction of social capital previously accumulated by Swedish society. Still “[i]n the beginning of the 1980s, 82 percent of Swedes (…) agreed with the statement ‘Claiming government benefits to which you are not entitled is never justifiable.’ (…) In the survey conducted between 2005 and 2008, only (…) 61 percent of Swedes believed that it was never right to claim benefits to which they were not entitled. (…) The survey conducted between 2010 and 2014 (…) found that (…) merely 55 percent of Swedes answered that it was never right to overuse benefits” [Sanandaji, 2016b, p. 131].

According to Sanandaji, the longevity of the Nordic countries’ inhabitants, including the Swedes, is not at all due to the existence of an extensive welfare state in these countries, of which the health care financing system is a part. Rather, it is the result of their genetic characteristics, a healthy diet, and, in general, a healthy lifestyle.

Justifying this thesis, Sanandaji indicates, among other things, that as early as 1960 Swedes lived 3.2 years longer than Americans (I mean the average life expectancy of a person aged 0), even though the size of the public sector in Sweden, as measured by the share of taxes in GDP, was then similar to that in the United States. Moreover, with the expansion of the public sector in Sweden (but not in the United States), today – writes Sanandaji in 2016 – this difference has decreased to 2.9 years, not increased [Sanandaji, 2016b, p. 32].

When we look at the changes in the composition of the group of 10 countries with the longest life expectancy in the world it turns out that Sweden, in 1960, which was ranked fourth in this ranking, in 2000 and 2013, was in 10th place, overtaken by countries with a relatively small public sector, such as Japan, Switzerland or Singapore [Sanandaji, 2016b, p. 34]. The conclusion seems obvious: an extensive social welfare system does not translate directly into high life expectancy. Other factors are also important, such as diet and lifestyle [Sanandaji, 2016, p. 35]. Sanandaji even suggests that an overly large welfare state can hamper the extension of the average life expectancy [Sanandaji, 2016b, p. 32].

Likewise, maintains Sanandaji, is with child mortality under five years old. Admittedly, in Sweden in 2015 it was almost twice as small as in the United States. However, if we compare the situation in Sweden with that of many other countries, it turns out that Sweden already boasted a very low child mortality rate (first place in the world!) in 1960 [Sanandaji, 2016b, p. 37]. Again, over the succeeding decades, as the welfare state expanded, Sweden gradually moved down in this ranking. Countries with a relatively small public sector, such as Japan, Singapore, and South Korea [Sanandaji, 2016b], were heading upwards.

According to Sanandaji, income inequality also decreased the most in Sweden in the 19th century, as well as in the first half of the 20th century, long before the establishment of the great Swedish welfare state. He quotes Swedish economists Jesper Roine and Daniel Waldenström: “We find that, starting from levels of inequality approximately equal to those in other Western countries at the time, the income share of the Swedish top decile drops sharply over the first eighty years of the twentieth century. Most of the decrease takes place before the expansion of the welfare state and by 1950 Swedish top income shares were already lower than in other countries” [Sanandaji, 2016b, p. 38; cf. Roine and Waldenström, 2008, p. 366].

In general, according to Sanandaji, the small income differentiation of the Swedes is more a result of the homogeneity of society than the actions of the welfare state. He lists the 10 countries with the least inequality (Slovenia, Denmark, Norway, Iceland, Slovakia, Czech Republic, Finland, Belgium, Sweden, and Austria) and points out that their common feature is not the great welfare state but the homogeneity of society [Sanandaji, 2016b, p. 40]. As he writes: “[a] homogeneous population means that the majority of citizens share the same culture. When this is the case, unsurprisingly, incomes are likely to be more similar than in countries with big differences in culture. In fact, we know that the large differences in culture in the United States is a main cause of high inequality” [Sanandaji, 2016b, p. 40]. Sanandaji bases his opinion on the results of Tom Hertz's research, which shows that in the United States the differences in the average income of different ethnic groups (e.g., African-Americans, Asian-Americans, white Americans) are very persistent [Sanandaji, 2016b, pp. 40, 200; cf. Hertz, 2008, pp. 415–417].

Sanandaji also adds that the small income inequalities in the Nordic countries are, among other things, the result of the innate egalitarianism of Nordic societies and not the actions of the welfare state [Sanandaji, 2016a, pp. 16–18; 73–76; cf. Sanandaji, 2016b, p. 39].

In general, according to Sanandaji, the US's free market offers its citizens much better conditions for development than the caring Nordic countries. For example, Sanandaji points out that the descendants of poor Swedish emigrants who came to the United States en masse in the 19th century, after more than a century in many respects (relatively high income, low unemployment, low poverty rate, and good education) are much better off today than both the rest of the Americans and the descendants of those Nordic citizens who did not decide to emigrate from their homeland years ago [Sanandaji, 2016a, pp. 79–82; cf. Sanandaji, 2016b, pp. 58–68]. Independently, according to Sanandaji, the fates of these Swedish immigrants to the United States show – once again – the decisive importance of Nordic culture (i.e., social capital) for economic development.

Therefore Sanandaji believes that the famous “American dream”

The American Dream represents the image of the United States as a country where everyone, through their abilities and work, regardless of their place of birth and social background can achieve success and wealth (cf. “The American dream.” Merriam-Webster.com Dictionary, Merriam-Webster, https://www.merriam-webster.com/dictionary/the%20American%20dream (accessed 19 December 2021).

Indeed, data on intergenerational income mobility shows that it is greater in the Nordic countries than in the United States. However, Sanandaji writes that this does not lead to the conclusion that the chances of social advancement and transition to another social class are greater in the Nordic countries than in the United States. The United States, until recently unlike the Nordic countries, has been an ethnically and culturally heterogeneous country. “When you think about it, it simply doesn’t make sense to compare the highly diverse American society with highly homogeneous Nordic societies” [Sanandaji, 2016b, p. 149]. More than half of the relationship between children's income and parents’ income in the United States is due to the persistence of differences in the average income of different ethnic groups [Sanandaji, 2016b ]. In this context once again Sanandaji refers to Hertz's research [Sanandaji, 2016b, cf. Hertz, 2008, pp. 415–417].

In his article from 2018 Matt Bruenig writes: “[f]or instance, Sanandaji argues that Nordic longevity is partially caused by healthy lifestyles, not just high-quality public health care (which Sanandaji nonetheless thinks is a good idea — ‘I am myself in favor of the public sector providing health care’). Characterizing this point as ‘debunking’ seems a bit much insofar as I’ve never seen anyone argue that the presence of public health care is the sole determinant of lifespans” [Bruenig, 2018].

Further, Sanandaji argues that after 1960, as the public sector in Sweden grew, the Swedish advantage over Americans in terms of life expectancy diminished (from 3.2 years in 1960 to 2.9 years before the publication of Debunking Utopia in 2016). Sanandaji even suggests that an overly large welfare state can hamper the extension of the average life expectancy.

However, OECD data for the same period say something different [OECDa; OECDb]. The Swedes’ advantage over the Americans in terms of life expectancy did not decrease, but grew and was 3.2 years in 1960, 3.2 years in 2013, 3.4 years in 2014, 3.6 in 2015, 3.7 years in 2016, 3.9 years in 2017 and 3.9 years in 2018. Moreover, between 2013 and 2018, in the United States, this was accompanied by a decrease (−4.8%) and in Sweden, the share of tax revenues in GDP has increased ((+2.5%). This contradicts the data Sanandaji refers to, as well as his entire argumentation.

Indeed, between 1960 and 2013, in terms of life expectancy, Sweden moved from the 4th to 10th position in the world and was overtaken by countries with a relatively small public sector in the economy (e.g., Switzerland). It was similar to child mortality. However, some of the countries that are ahead of Sweden (e.g., South Korea, Japan, Singapore) differ significantly in e.g., social discipline, work ethos, lifestyle, including eating habits, from Western countries. This casts doubt on Sanandaji's suggestion that a small public sector is the source of these countries’ successes. In addition, the public sector in yet other countries, which in the same period of time overtook Sweden, in 2013 was not small at all, but very large (larger than in Sweden) (e.g., Italy and France).

In 2013, the ratio of tax revenue to GDP was 42.86%, 45.37%, and 44.05% in Sweden, France, and Italy.

Finally, Sanandaji overlooks the fundamental role of the Swedish state in improving the health of Swedish society in the 18th, 19th, and then in the first half of the 20th century. Here I am thinking of, e.g., activities of the Collegium Medicum, then Collegium of Health (Sw. Sundhetskollegium), and the National Swedish Health Council (Sw. Medicinalstyrelsen) and the promotion of hygiene, construction of sanitation, development of hospital networks. The same applies to the Swedish Social Democrats’ pro-health policy that has been pursued practically since the party was founded in 1889 [Porter, 2005, pp. 61, 96–97, 183–184].

Sanandaji's analysis of the reasons for the low-income inequality in Sweden is superficial. For example, his belief in the historically conditioned egalitarianism of the Swedes is contradicted by (otherwise quoted by Sanandaji on p. 38 Debunking Utopia) Jesper Roine and Daniel Waldenström's observation that at the beginning of the 20th century the level of income inequality in Sweden was not lower than in other countries).

Similarly, in defending the thesis that the small income disparities in Sweden are the result of the cultural homogeneity of society rather than the actions of the welfare state, Sanandaji uses the example of the group of countries where income inequalities are the smallest. He writes that their common feature is not a large welfare state, but the uniformity of society.

However, apart from the Nordic countries, three of the other five countries are former “real socialist” countries (Czech Republic, Slovakia, and Slovenia). It can be assumed that the small inequalities are, to a large extent, caused by their history. In addition, in Slovenia, the public sector was larger than the OECD average in 2015. This causes Sanandaji's entire argumentation to boil down to pointing out two countries with a homogeneous culture and small inequalities (Austria and Belgium). It is difficult to see this as a strong argument in this discussion, especially since Austria and Belgium are countries with a very large public sector; in 2015, the share of the country's tax revenue in GDP was 43.04% and 44.73%, respectively (sixth and third place in the OECD) [OECDc]. Nor does Sanandaji's example of the United States as a country where large cultural differences are the main cause of large inequalities convince. The cause of large inequalities in the United States can be considered, as Joseph Stiglitz does, for example, to be because of the lack of effective economic policies to reduce inequalities in the United States [cf. too The Economist, 2019].

In addition, Sanandaji's examples can be contrasted with the counterexamples. Criticizing Sanandaji's view that the cause of income inequality is cultural heterogeneity in society, Ingvild Reymert points out, for example, that over the course of half a century, the United States and Japan have changed places in international rankings of income diversification. In the 1950s, income differences were large in Japan and small in the United States, and today the opposite is true. The explanation for this event is economic policy, not culture [Reymert, 2016].

Matt Bruenig [Bruenig, 2018] has given a detailed analysis of the causes of minor inequalities in Sweden and other Nordic countries. According to Bruenig, Sanandaji's arguments are not convincing.

Firstly, Sanandaji refers to income before taxes and benefits. Meanwhile, as Breunig points out, “Needless to say, an income concept that intentionally ignores the welfare state is not going to tell you much about the effect of the welfare state. For that, you need to measure disposable income, not pretax income (Bruenig's italics – B.Cz.)”. Secondly, the data referred to by Sanandaji are either incomplete or (this is the case most often) refer to the upper-income decile in Sweden and other Nordic societies at most. As a result, the picture of the evolution of inequalities in the Nordic countries presented by Sanandaji is simply incomplete and thus misleading. “There is a whole 90 percent of people below the top 10 percent, and if you want to talk about inequality, you need to break down the distribution of their incomes as well. This is especially true when you are talking about poverty, because poverty is about how much income the bottom gets relative to the middle. It has nothing to do with top incomes. Yet Sanandaji cites this evidence as proving ‘low levels of poverty’ precede the welfare state.”

In general, it is clear to Bruenig that it is not the cultural homogeneity of Swedes that is the reason for the reduction of income inequality in Sweden in the first half of the 20th century. The results of Jesper Roine and Daniel Waldenström's research are unambiguous. This is due to a unique sequence of historical events (wars and economic crises), as well as the policy of the emerging social democratic welfare state (tax increases) and the actions of Swedish trade unions (a relative increase in wages for work and a decrease in the return on invested capital).

In particular, Bruenig writes that the reason for Sanandaji's indicated a decline in the income of the Swedish upper decile in the first half of the 20th century was – mainly – the dramatic decline in income from capital and doing business of the upper 1% of the most affluent Swedes. As established by Roine and Waldenström in 2006 and referred to by both Sanandaji and Bruenig, this reason is responsible for 78% of the fall in the revenue share of the upper percentage of Swedes in GDP between 1912 and 1935 and for 57% of that fall between 1935 and 1951.

Bruenig points out two reasons for this decline in income from capital. This concerns the decrease in the share of income of capital in the production value, as well as the decrease in the percentage of Swedes with the highest income in capital ownership (national wealth ownership) in Sweden. Bruenig writes that, firstly, before 1920, about 35–40% of the income included in GDP in Sweden was capital income, and in 1950 this share fell to about 25%. Secondly, at its peak, between 1910 and 1920, the upper (in terms of income) 1% of Swedes owned >0% of the national wealth in Sweden. By 1950, this percentage had fallen to about 25%.

According to Bruenig, “[w]hen you put the two causes together, you get a pretty clear story about what happened to the pretax incomes of Sweden's top 1 percent before 1950: they lost over half of their share of the national capital, and the capital they had left started receiving a much lower rate of return. These two changes dramatically reduced the amount of capital income the top 1 percent received, and thereby dramatically reduced their share of total national income.”

Bruenig writes that the decrease in the share of the upper percentage of Swedes in capital ownership was caused by the Great Depression, including, for example, the collapse of the empire of the Swedish industrialist Ivar Krueger and war damage. Admittedly, according to Roine and Waldenström, the war damage to Swedes’ property was of little concern, as Sweden did not participate in either world war [cf. Roine and Waldenström, 2008, p. 383]. Bruenig is of a different opinion, however, arguing that the data referred to by Roine and Waldenström is incomplete and that “affluent Swedes likely owned assets outside of their own country.”

Roine and Waldenström reiterated their opinion in their 2015 work [Roine, Waldenström, 2015, p. 555]. In their view, the decline in the income of the richest from capital was caused not only by a progressive decline in the share of remuneration of capital in GDP but also by the economic policy of the state, especially the strong increase in the highest tax rates. Increases in income and property taxes discouraged the richest from working and accumulating capital, which resulted in a decrease in their income.

Bruenig concludes his analysis by pointing out that, according to Roine and Waldenström's findings, after 1950, the decline in the share of the upper percentage of the richest Swedes in national wealth became permanent. The reason for its consolidation was a further increase in taxation, i.e., the policy of the welfare state, which – by raising taxes – prevented the key, richest percentage of Swedish society from rebuilding its damaged wealth. The result was, among other things, a sharp increase in public property. In 1950, the Swedish state held 14% of the national wealth, and in 1977 its share of the wealth increased to 43%.

Sanandaji argues that the chances of the “American Dream” materializing are greater in the United States today than in Sweden and other Nordic countries. He points out, among other things, that the average income of the descendants of Nordic immigrants in the United States is higher than the average income of the descendants of those Nordic citizens who decided to stay in their homeland many years ago.

Einar Lie and Jo Thori Lind analyzed Sanandaji's argumentation in detail. [Lie and Lind, 2016]. They start by reminding us that comparing the situation of the descendants of emigrants with the situation of the descendants of those who have remained in the country involves many problems that Sanandaji ignores. In fact, for example, immigrants coming to the United States from Nordic countries were not a representative sample of Nordic societies. However, even if we reject such objections and accept Sanandaji's point of departure, the conclusions we reach by following in his footsteps turn out to be different from his.

Lie and Lind point out that the descendants of almost all immigrants (regardless of their nationality) who came to the United States years ago are now earning a higher income than the inhabitants of the countries they left, for many reasons. These include abundant natural resources, a large supply of skilled workers, the size of the internal market, and the mobility of both labor and capital in the United States. All these reasons have led the United States, long before the emergence of modern welfare states, to overtake European countries, including the Nordic countries, in terms of per capita GDP.

Next, Lie and Lind argue, the relative, not absolute, size of the excess of income of the descendants of Nordic immigrants arriving in the United States over that of the current residents of those countries is important. Namely, it is important whether this surplus is higher or lower than the corresponding surplus for descendants of immigrants coming to the United States from other countries.

According to Sanandaji's argumentation, this surplus of income of descendants of immigrants from the Nordic countries should be high compared to the similar surplus of income of descendants of immigrants from other countries. First of all, their social capital should provide them with a particularly large increase in income over more than a century compared to the descendants of immigrants from other countries. Secondly, the increase in the income of the residents of the Nordic countries compared to the increase in the income of the residents of other countries should be relatively small. After all, according to Sanandaji, income growth in the Nordic countries is held back by the overdeveloped Nordic welfare states.

In fact, it turns out that in the case of descendants of immigrants from the Nordic countries this surplus is relatively small compared to a similar surplus of descendants of immigrants from other countries. This conclusion is based on two premises. First, the relevant statistics show that the level of GDP per capita in the Nordic countries is, compared to other countries (including European countries), very high. Second, the analysis of the American Bureau of the Census for 2009 reveals, Lie and Lind write, that the descendants of immigrants from Nordic countries have incomes similar (or slightly lower) to those of immigrants from other European countries.

In particular, income higher than that of the descendants of immigrants from the Nordic countries is achieved, among others, by the descendants of immigrants from Luxembourg, Austria, Russia and the Baltic States, the United Kingdom (including the English, Scots, and Welsh), as well as a large group of people declaring themselves to be descendants of “immigrants from Europe.”

Nordic social capital is therefore not sufficient for the descendants of Nordic immigrants who arrived in the United States more than a century ago to gain an income advantage in America over the descendants of other immigrants. And yet the people of the Nordic countries, against the background of Europe and the world, have particularly high incomes. It is not this capital, therefore, that is probably the main reason for the relatively high incomes of the Nordic countries’ inhabitants compared to those of other countries. This is probably due to the way Nordic economies are organized and, possibly, the specific economic policies of the Nordic countries.

Let's sum up. The data referred to and the reasoning behind it justify the conclusion that, under comparable conditions, the inhabitants of the Nordic countries with their social capital are more successful in the Nordic countries than in the United States. In other words, the social and economic development support institutions created in the Nordic countries, as well as Nordic economic policy, provide a better environment for development in those societies than the institutions and economic policy in the United States.

In this way, Lie and Lind at one fell swoop contradict Sanandaji's opinion that the successes of the Nordic countries are primarily due to their social capital and that in the Nordic countries excessive state intervention (an overly tax-funded, over-expanded “welfare state”) has impeded economic development. If these views were true, for the descendants of immigrants from the Nordic countries, the surplus of their income over that of Nordic residents would turn out to be higher than that of immigrants from other countries, in particular those whose economies are more liberal than those of the Nordic countries. By the way, it turns out that the “American Dream” is closer to becoming reality in the Nordic countries rather than in the United States.

Sanandaji is right when he writes that “[s]ome far-reaching conclusions have been based on the observation that parental incomes have a stronger link to the incomes of their children in the United States than in the Nordics [Sanandaji 2016b, p. 149]”. An example is Miles Corak, who thus summarizes the results of his research on income mobility differences in the world: “Many Americans may hold the belief that hard work is what it takes to get ahead, but in actual fact the playing field is a good deal stickier than it appears. Family background, not just individual effort and hard work, is importantly related to one's position in the economic and social hierarchy. (…) In fact, children are much more likely as adults to end up in the same place on the income and status ladder as their parents in the United States than in most other countries” [Corak, 2016, p. 2].

Sanandaji believes that the relatively low intergenerational income mobility in the United States does not mean that social advancement is more difficult in this country than in the Nordic countries. The United States is much more heterogeneous ethnically and culturally than the Nordic countries. Sanandaji stresses that the average income of Americans of African, Asian, Latin American, and white origin varies greatly. As a result, more than half of the relationship between children's income and parents’ income is due to the persistence of differences in the average income of these ethnic groups, which – according to Sanandaji – makes comparisons impossible.

In short, Sanandaji points to the relative persistence of income differences between different ethnic groups in the United States. He underlines that, according to Hertz, these differences explain about half of the small intergenerational mobility in the United States. In his opinion, these circumstances somehow invalidate (justify?) the conclusions of the high level of intergenerational income elasticity in America, e.g., Miles Corak's conclusions about the chances of social advancement in America.

Commenting on Sanandaji's opinion, Clara Hendrickson notes: “Unfortunately, he (Sanandaji – B.Cz.) doesn’t adequately explain why using race and ethnicity to compare relative social mobility offers a more accurate measure than using social class. Instead, he assumes that homogeneity is the bedrock of Nordic prosperity” [Hendrickson, 2017]. Indeed, for a consistent advocate of “equal opportunities,” the relative persistence of income differences between different social groups and the accompanying low social mobility deserve a negative assessment, regardless of whether these groups have been distinguished based on a class criterion or based on ethnic or cultural criteria.

For example, the philosopher John Rawls calls equality of opportunity a situation in which people with the same talents and abilities, and those equally willing to use those talents and abilities, have an equal chance of success, regardless of circumstances such as class, race, and gender, (my italics – B.Cz.) [Mason, 2020]).

Joseph Stiglitz similarly agrees. He enumerates the barriers to the social advancement of people of colour in the United States. “Residential and educational segregation leads to less opportunity, and employment discrimination means that getting a job is more difficult for people of color. This structural discrimination creates large wealth gaps between whites and other population groups - inequalities that transmit down through generations from parents to children. This is especially troubling given that people of color make up a majority of America's future workforce” [Stiglitz, 2016, Kindle Locations 1122–1126].